URN:NBN:fi-fe20031090 The Imperial mints of Philip the Arab Jyrki Muona Zoological Museum, JP 17. FIN-00014 University of Helsinki [email protected] Introduction This paper is a by-product of research aiming at finding out, whether the methodology of phylogenetic systematics requires an underlying hypothesis of evolution in order to be successful. Many, if not most, systematists believe that cladistic analyses require a general hypothesis of an evolutionary process (e.g. Farris, 1983, Kluge, 2001), but others claim that hierarchical character distributions suffice (e.g. Brower, 2000). These arguments are based on theoretical premises, as they should. The question could be approached from a practical point of view as well, however. Not only living things, but also any set of hierarchically organized items defined by characters could be analysed. Such a study might indicate, whether or not the method used in phylogenetic systematics operates in the proven absence of real ancestor-descendant relationships. The true phylogenies of living organisms cannot be known as they are always inferred from data, but minting sequences of coins can be known. The results of the coin analyses can be compared with external historical information, which either corroborates or contradicts the outcome. This situation never holds in biology. The objects studied need to possess characters showing variation, just as in biological systematics observable characters are needed for defining taxa. “Overall similarity”, a term with much the same meaning in biological systematics as the term “style” has in numismatics, is known to be misleading in many instances. Instead of overall similarity, modern biologists search for unique, derived characters uniting taxa. An analogy illuminating this point may be in place here. The change from "GALBA" to "OTHO" in the emperor's name on coins can be seen to be analogous to a change in the shape of a bone between a lizard and a bird. Both hypotheses are based on observed differences in features, which the observer believes to be the same thing: "the emperor's name" in coins and "the hip-bone" in animals. All changes are not equal, however. The words "CAESAR" and "AVGVSTVS" cannot be used as characters solving the relationships within a set of three coins, two of them minted under Otho and one under Nero, as both are "inherited" from earlier coins. On the other hand, the word "OTHO" certainly units the two right ones and shows their actual relationship. The reason the word "OTHO" could be assumed to work and the others not, is the fact that although "CAESAR" and "AVGVSTVS" define a group of coins of certain age and type, this group include all the three coins under study and the words in fact define a much larger group of coins than the ones we are interested in in this particular case. In order to find the correct result, one needs characters defining smaller groups within the study group. My first, preliminary attempt dealt with the Rome mint coins of Marcus Salvius Otho (Muona, 2002b). The results suggested that although coin types could be defined with discreet characters (self-evident, of course), the minting sequence of the types was not necessarily recoverable with the method employed. In order to get further insight, I set out to study the antoniniani of Philip the Arab and his family. The starting point was simple: could one find discreet characteristics that would place the antoniniani of Philip the Arab and his family to a particular mint? During the work I observed features previously ignored when classifying these coins (or at least unreported), including the variation in the structure of the cuirass and the letters in the legends. As these characters provided useful hints for the separation of the mints, a note reporting the observations seemed to be in place. Identifying the mint a Roman coin came from can be difficult. Most early Imperial coins lacked direct evidence of origin. This holds for the reign of Philip the Arab as well. A study of his coins is complicated further by the paucity of contemporary written evidence. Philip employed many mints, both Imperial and Provincial, to produce a diverse and complex coinage. An interesting peculiarity of his coinage was the minting of “Imperial” coins in the provinces (at least in Antioch) and “Provincial” coins in Rome (at least for Antioch) (Baldus, 1969, also Prieur & Prieur, 2000). Some of the tetradrachms had the ethnic clearly marked (“ANTIOXIA/SC”). They were stylistically identical with an extensive series of antoniniani (from Antioch mint II sensu Mattingly), and both types must have been minted in Antioch. Another group of tetradrachms was characterized by the legend "MON VRB". The “MON(ETA) VRB(S)” is assumed to refer to the city of Rome. These coins were stylistically identical with the Rome mint antoniniani (but had legends in Greek) and they must have been minted in Rome for use in the East. According to RIC IV (Mattingly, 1949), Philip minted antoniniani in Rome and Antioch. Two different types of Antioch coins were listed in RIC, referred to as Antioch mints "I" and "II". According to Òvari (1984) and Clay (1997), three or four different mints existed (see also Muona 2002a). In addition to the mints in Rome and Antioch (= Antioch "II" sensu RIC), one mint in Viminacium and an "Unknown mint" (= Antioch "I" sensu RIC) produced coins for Philip. I will call these four mints the Rome, the Antioch, the Viminacium and the Unknown mint throughout this paper. A small series of coins minted for Philip II (RIC 213, 214) and Otacilia Severa (RIC 127, 128) were tentatively placed to Antioch in RIC. The origin of these coins was complicated by the fact that no issues of Philip I was associated with them. As pointed out by Òvari (1989) (see also Muona, 2002a), the Antioch mint coins of the Philips were very poorly dealt with in RIC. The case of Otacilia Severa seems to have been even worse. Le Gentilhomme (1946) placed the Otacilia coins RIC regarded as definitely Antiochian, RIC 132 to 134, to the Unknown mint. Thus only coins that RIC called “hybrids” or having “reverses of Philip I” were left to represent her in Antioch sensu stricto. Material The number of different Philip I and II dies seen from each putative mint forms the data for the cuirass study. The true number of dies is very high and my small sample is not intended to be proportionately representative as far as the reverse types are concerned. Only after a proper die-analysis could one attempt this and it is doubtful that such a huge coinage as Philip the Arab's will ever be so analysed. However, as dies were searched from quite randomly from available coins, larger issues are better represented than smaller ones. In several cases the characters were studied from many coins of the same die. The variation in strike and wear caused problems of interpretation and many coins seen could not be included in the data. The lettering was studied from coins of Otacilia Severa as well. The coin types studied are listed in table 1. The cuirass Antioch mint The Antioch mint produced an extraordinary rich sequence of antoniniani under Philip (Òvari, 1989). Six different bust types are known, all of them cuirassed, four also draped. The commonest type, as in Rome, shows draped and cuirassed bust right, seen from behind. Two other right-facing types exist, but they are much rarer. On them the bust is seen from front and either cuirassed or cuirassed and draped. These two unusual types show nicely the detailed shoulder structure of the cuirass, with a row of hanging straps and a series of knots (Figs. 1-2). This Antioch style cuirass is clearly visible in the commonest bust type (Figs. 3-4), which was engraved in more detail than that of the Rome mint coins. There are five features to look for (Fig. 3): (1) the knots at the tips of the straps (a), (2) the knots at the edge of the actual cuirass, under the fibula (b) (3) a transverse ridge depicting the cuirass edge (c), (4) a prominent dot left of the main folding of the paludamentum (d - often there is a dot on the side of the shoulder as well), (5) the number of straps and the width of the opening in drapery. Typically the Antioch mint coins show both types of knots and the dot, have a fairly narrow drapery opening and from 4 to 7 straps, the average being 4.44 (table 2). Usually the transverse ridge is present. The coins of Philip II show more variation than those of his father. The dot is missing from about a third of the dies and the knots or the ridge can be absent - two dies showed none of these features. The average number of straps is higher than in the coins of Philip I, 5.53 (table 2). Rome mint The Rome mint dies of the antoniniani of Philip I show a fairly uniformly engraved cuirass (Figs. 5-7). The cuirass is seen through a narrow opening of the paludamentum under the fibula, just as in the Antioch coins. The straps are shown as bent ridges and there is no transverse basal ridge indicating the edge of the actual cuirass. The number of straps varies from 4 to 7, the average number being slightly higher than in the Antioch mint Philip I coins, 4.95 (table 2). The coins of Philip II as augustus are identical to those of his father with respect to the cuirass details. There is no difference between the coins of the father and the son in the number of the cuirass straps.